Project Summary This project addresses the urgent need of a comprehensive model to further our understanding of the complex and dynamic role of various risk factors in early childhood caries (ECC) development across the child's age continuum. Although the risk factors (biological, behavioral, socio-economic, psychological and environmental) of ECC are relatively well known, their effects and their discriminating power along the disease pathological process remain poorly understood. A popular approach for evaluating these effects has been to use models that impose homogeneity along the life course of the disease, or to rely on cross-sectional data. This approach is useful but has limitations. Some risk factors may be more important in very early life (eg., low birth weight and other biological conditions), but less relevant in older ages as other risk factors come into play and the children become more vulnerable to caries. A unique study that provides detailed longitudinal information to untangle the cascade of relevant risk factors along the life course of ECC is the Detroit Dental Health Project. This is a longitudinal study designed to understand the oral health of low-income African-American children (0-5 years) and their main caregivers (14+ years) residing in the city of Detroit, Michigan. Among other information, this study comprises longitudinal investigations involving various risk factors and mouth-level caries indices which can help to delineate the natural trajectories of ECC in an unfortunate setting of limited clinical intervention. Analyzing ECC data from the Detroit study is inherently complex because of the large number of ECC risk factors that are potentially correlated; complex relationships between these risk factors and caries along the child's age continuum; the longitudinal (within-child) association of caries indices; and a sizeable frequency of zero-valued caries indices. To our best knowledge, we are not aware of any statistical method that addresses these nontrivial complications that are very specific to ECC indices and the Detroit data. We propose to fill this gap in the literature by formulating a less restrictive semiparametric model coupled with a variable selection method to understand the life course of ECC among inner-city African-American children. Formulating and implementing such a model is not trivial owing to the complexity of the theoretical arguments and the computational burden. Following the completion of this project, we will acquire evidenced-based knowledge that would be helpful in tailoring a dental caries prevention policy for low-income children during the first five years of life.